Gareth DaviesSencillo Research Ltd; Kwiziq Ltd; Beat Fitness Games Ltd
Gareth Davies
PhD (Experimental Physics)
About
5
Publications
1,608
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
28
Citations
Introduction
Current research focuses on ML techniques in NLP and time series inverse problems.
Since March, I've been working pro bono on COVID-19. I developed a new method for outbreak severity estimation; and proved that vitamin D deficiency is the dominant risk factor in poor COVID-19 outcomes.
Spare time: theoretical physics research (quantum gravitation, cosmology & black holes).
Physics BSc, Medical Physics PhD (Imperial). Novel ways to solve the EM inverse problem for MCG/MEG.
Education
July 1991 - June 1994
Independent Researcher
Field of study
- Medical Physics
August 1988 - June 1991
Independent Researcher
Field of study
- Physics
Publications
Publications (5)
UK Biobank analyses concluded that COVID-19 risk, particularly higher risk in ethnic minorities, was not explained by vitamin D.[1–10] Hastie, et al. dismissed previous critiques, responding that their analyses were "…as powerful as any to date".[8] However, the reported statistical significance and high precision are illusory; these papers used un...
Serious Statistical Flaws in Hastie, et al., Vitamin D concentrations and COVID-19 infection in UK Biobank Analysis
Gareth Davies (https://orcid.org/0000-0002-8901-0680)
Richard B. Mazess (https://orcid.org/0000-0001-6647-4311)
Linda L. Benskin (https://orcid.org/0000-0002-6096-3971)
UK Biobank analyses concluded that COVID-19 risk, particularly...
We demonstrate that studies investigating Vitamin D concentrations and COVID-19 using Biobank data by Hastie, et al., lacked power, suffered low precision and high bias, used flawed models, and contained many serious statistical errors. These can be summarised into three areas: 1. Mislabelled data artificially inflated the control set, creating an...
We analyse global data for COVID-19 deaths and recoveries and show that outbreak severity displays a striking latitude relationship with a northern hemisphere bias. Transmission rates can be explained by seasonal weather conditions, but this does not account for observed variations in fatality rates. Many factors point to Vitamin D as a candidate e...
Determining the severity of a novel pathogen in the early stages is difficult in the absence of reliable data. The pattern of outbreaks for COVID-19 across the globe have differed markedly above and below +40N latitudes, suggesting very different levels of severity, but countries worldwide have implemented the same lockdown strategies. Existing met...
